vector-db-ingest — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited vector-db-ingest (Agent Skill) and scored it 87/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 3 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 3 flagged
The text {match} tells the agent to skip the normal "ask the user first" gate. Used adversarially it removes the human-in-the-loop check before destructive or sensitive actions, turning a normally-gated agent into a fire-and-forget executor.
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
This skill requires the chromadb and langchain packages defined in the plugin root.
You ingest (index) repository files into the ChromaDB vector store so they can be semantically searched. You build or update the parent-child chunk structure that query.py searches against.
High-Performance Mode: This skill uses a configurable batch processing engine (default 1,000 files) defined in .agent/learning/vector_profiles.json.
If vector_profiles.json is missing, run the init skill first:
python ./scripts/init.pyThis plugin defaults to In-Process mode for zero-latency direct disk access. No background server is required unless explicitly configured in the profile.
Note: The --profile flag is mandatory to load the correct manifest and batch settings.
python ./scripts/ingest.py --profile wiki --fullpython ./scripts/ingest.py --profile wiki --since 24python ./scripts/ingest.py --profile wiki --file path/to/file.md
python ./scripts/ingest.py --profile wiki --folder path/to/folderRun a quick semantic search to confirm the new content is retrievable:
python ./scripts/query.py "search query" --profile wiki --limit 3--profile to ensure the correct batch size and manifest are used.~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.